Robustness of steel joints with stainless steel bolts in ﬁre · Keywords Steel joints Stainless...

ORIGINAL RESEARCH

Robustness of steel joints with stainless steel bolts in fire

N. Satheeskumar • J. B. Davison

Received: 22 October 2014 / Accepted: 5 November 2014 / Published online: 12 November 2014

� The Author(s) 2014. This article is published with open access at Springerlink.com

Abstract The robustness of steel joints in fire is important

for steel building structures because of the need to prevent

progressive collapse. Stainless steel is widely used in

building construction mainly because of its corrosion resis-

tance, but it also possesses improved fire resistance com-

pared with conventional non-alloy, fine grain structural

steels. Extensive research performed on the robustness of

steel joints in fire has revealed that failure at elevated tem-

perature may be controlled by bolt shear for fin plate and web

cleat connections. Hence, this study focussed on the use of

stainless steel in experimental tests conducted on fin plate

and web cleat connections at high temperatures. In addition,

this study investigated the use of a component-based model

to predict connection performance at elevated temperature.

Keywords Steel joints � Stainless steel joint � Web cleat

connection � Fin plate connection � Robustness

Introduction

Non-residential multi-storey buildings often use a steel-

framed structure and steel-concrete composite floor and

beams because of speed of construction, low cost, lighter

weight and low maintenance cost when compared to a con-

crete structure. The performance of steel-framed structures

in fire, although now well understood and accounted for in

design and construction by a range of approaches, remains a

cause of concern for designers in some parts of the world.

The damaging effect of fire and resulting collapse of a

number of buildings at the World Trade Center (New York)

in 2001 focussed attention on the need to ensure robustness in

steel-framed buildings under extreme conditions.

Current design codes generally consider that steel con-

nections will be heated more slowly than beams or columns in

fire situations, therefore reducing the likelihood that they will

be the critical components in fire safety design (Al-Jabri et al.

2008). In the UK, connections in structural steel are designed

to transfer shear (and sometimes moment) at ambient tem-

perature and also resist an axial force, known as a tying force,

under accidental loading conditions (Way 2011). In fire, the

joints experience additional compressive or tensile forces due

to restraint to thermal expansion or to catenary action arising

from large deflections (Yu et al. 2009). Connection failures

could lead to progressive collapse of a building. The term of

progressive collapse is ‘‘the spread of an initial local failure

from element to element, eventually resulting in the collapse

of an entire structure or a disproportionately large part of it.’’

(UFC 2009). Therefore, it is essential that the structural ele-

ments and the connections between them have the ability to

resist a degree of damage in order to prevent progressive

collapse. This inherent characteristic of a well-designed

structure is referred to as robustness.

The term Robustness is defined as the ‘‘ability of a

structure to withstand events like fire, explosions, impact or

the consequences of human error, without being damaged

to an extent disproportionate to the original cause’’ (CEN

2002). The ‘‘Tying force’’ approach is one of the design

methods used to provide a measure of resistance to

N. Satheeskumar (&)

School of Engineering and Physical Sciences, James Cook

University, Angus Smith Drive, Douglas, Townsville,

QLD 4811, Australia

e-mail: [email protected]

J. B. Davison

Department of Civil and Structural Engineering, The University

of Sheffield, Sir Frederic Mappin Building, Mappin Street,

Sheffield S1 3JD, UK

e-mail: [email protected]

123

Int J Adv Struct Eng (2014) 6:161–168

DOI 10.1007/s40091-014-0075-0

progressive collapse. The term can be explained as ‘‘tying a

steel frame horizontally and vertically to increase its

structural continuity and create a structure with a high level

of robustness’’ (Yu et al. 2009). The ability to resist a tying

force is called tying resistance. The tying force generated

in a structural member (beam, column) must be conveyed

through the connection.

Research reported by Yu et al. (2009, 2009) considered

the elevated temperature robustness of fin plate and web

cleat connections. These studies highlighted the important

role played by the bolts and demonstrated that in many

cases, bolt failure became critical at elevated temperatures.

Stainless steel is widely used as an alternative to mild

steel for secondary steel work and fittings in building

construction because of its corrosion resistance, ease of

maintenance and aesthetic appearance. Stainless steel also

offers improved fire resistance, although this is rarely a

reason for choosing the material in structural applications.

However, this study sought to investigate the benefits of

replacing ordinary structural bolts with stainless ones in

joints subjected to high temperatures.

In this investigation, the test specimens were assembled

using stainless bolts in a fin plate and web cleat connections

which were identical to those tested by Yu et al. The

experimental test results are compared with the results and

conclusions of the earlier experiments by Yu (Yu et al. 2009,

2009) and also with component-based model analysis.

Experimental set-ups

The tests were carried out on typical beam-to-column

connections at a constant elevated temperature in an

electrically heated furnace. The temperature distribu-

tion around the joint was measured using thermocou-

ples which are located at the beam flange, beam web,

column flange, and fin plate and web cleat angles. The

specimens were heated to a specified temperature (550

or 650 �C) and tested at that constant temperature. The

load was applied through three, linked, strain-gauged

Macalloy bars so the load applied through the loading

bar could be determined as shown in Fig. 1. The

changes in inclination of these three bars were recorded

using three angular transducers and a digital camera.

The deformation of the specimen was recorded using

digital cameras and then determined by using image

recognition software. Figure 1 also shows the load

angle (a) between the furnace bar and the beam axis

determined the shear to tensile force ratio applied to the

connection.

A 254 9 89 kg/m universal column section in grade S

355 was used for the column and a 305 9 165 9 40 kg/m

universal beam section in grade S 275 was used for the

beam; these sections were used in all the experimental

tests. A 20 mm thick ceramic fibre blanket was wrapped

around the beam and column but not around the connection

zone (i.e. fin plate, web cleat angle, column flange, and

bolts and beam web). The connection zones are exposed to

heat throughout the testing.

Details of fin plate connection

Figure 2 shows the geometry of the fin plate connection

test specimen. The fin plate was 10 mm thick mild steel

and all the bolts were 20 mm diameter austenitic

Fig. 1 The experimental test

arrangement

162 Int J Adv Struct Eng (2014) 6:161–168

123

stainless steel of property class 80. The specimen was

tested at a temperature 550 �C and load angle (a, see

Fig. 1) of 55�.

Details of web cleat connection

The web cleat connection consisted of double angle

L90 9 90 9 8 sections in S 275 as shown in Fig. 3. Three

20-mm diameter austenitic stainless steel bolts in property

class 80 were used through the beam web and six 20-mm

diameter ordinary bolts in grade 8.8 were connected to the

column flange (the reason that grade 8.8 bolts were used in

the flange is because the failure mode in earlier tests (Yu

et al. 2009) was in the bolts in the beam web, that is, the

column flange bolts were not critical). The test was con-

ducted at a temperature of 650 �C and the load angle (a,

see Fig. 1) of 55o.

Component-based model assembly

Component-based model assembly for fin plate connection

The component-based model for the fin plate connection is

constructed as an assembly of spring elements as shown in

Fig. 4. This model uses a component-based model

approach developed by Sarraj et al. (2007) for a single line

of bolts in the connection. In the current analysis, a double

line of bolts was used in the connection. The active joint

components are the fin plate in bearing, beam web in

bearing, bolts in single shear, beam web-to-fin plate in

friction and weld in tension. This model was simplified via

a series of lap joints attached to each other in parallel

(Fig. 4). Figure 5 illustrates the lap joints of the fin plate

connection under tying force. All active joint components

are represented by a spring and the force–displacement

relationship of each was calculated according to Sarraj

Fig. 2 Fin plate connection details

Fig. 3 Web cleat connection

details

Int J Adv Struct Eng (2014) 6:161–168 163

123

et al. (2007) component-based model analysis. These cal-

culated force–displacement relationships were introduced

into the general FEM program ABAQUS simply as prop-

erties of spring elements.

Component-based model assembly for web cleat

connection

A component-based model analysis that was developed by

Yu et al. (2009) was used in this study. Yu’s research

identifies four active components in the typical web cleat

connection: bolts in tension, web cleats in bending, bolts in

shear and beam web in bearing. These active components

are assembled as a four-spring system in each series of bolt

rows as shown in Fig. 6. Yu’s research also investigated

the non-linear load and displacement response of web

angles under tensile loading. The current study has used Yu

et al. (2009) component-based model. The material prop-

erties and the force–displacement characteristics of the

bolts in double shear were modified to account for the use

of stainless steel. The force–displacement relationships of

each spring were calculated and introduced into the general

FEM program ABAQUS simply as properties of spring

elements.

Experimental results

The experimental results for the fin plate connection and

web cleat connection tests have been summarized in

Fig. 4 Component-based

model for fin plate connection

Fig. 5 Lap joint of fin plate

component model


123

Figs. 7 and 8 including the variations of the connection

rotation and force. Figure 7 shows the force–rotation

relationships for the fin plate connection at 550 oC and load

angle of 55o. It shows the maximum failure load of the fin

plate connection at 550 oC is 118 kN at a connection

rotation of 6o.

Figure 8 shows the variation of applied force versus

connection rotation for the web cleat connection at 650 �Cand load angle of 55o. It shows that the maximum failure

load of the web cleat connection at 650 �C is 28 kN at a

connection rotation of 10o.

Comparison of experimental tests results

Experimental test failure modes and force–rotation rela-

tionships and component-based model results of connec-

tions are compared with previous research results. Figure 9

illustrates the comparison of the force–rotation relationship

of the component-based model with the experimental test

results of the fin plate connection. It shows the maximum

failure load of the component-based model is 130 kN at a

connection rotation of 7.3o and the component-based sim-

ple model gives a reasonable prediction of the response of

the fin plate connection in fire. The robustness of the

connection, as defined by its tying resistance, is specified to

be a minimum of 75 kN (BSI 1990). At 550 �C, the fin

plate connection failure happened at a failure load that was

less than 75 kN in Yu’s test (Yu et al. 2009). In the current

experiment with stainless steel bolts, the connection failure

load was greater than the minimum tying force and

approximately two times higher than ordinary steel bolt fin

plate connection at elevated temperature.

Fig. 6 Component-based

model for web cleat connection

0

20

40

60

80

100

120

140

0 2 4 6 8 10 12

App

lied

Loa

d (k

N)

Connection Rotation (°)

Fig. 7 The force–rotation relationships for fin plate connection at

550 �C

0

5

10

15

20

25

30

35

0 2 4 6 8 10 12

App

lied

Loa

d (k

N)

Connection Rotation (°)

Fig. 8 The force–rotation relationships for web cleat connection at

650 �C


123

The failure mode in this study was weld fracture. Nor-

mally, weld fracture is not critical for a fin plate connec-

tion. The reduction in weld strength at high temperature

was proportionally greater than that in the stainless steel

bolts; thus the mode of failure transferred from plate

bearing at room temperature to weld fracture at high

temperature.

Figure 10 shows the comparison of force–rotation rela-

tionship of the component-based model with the experi-

mental test results of the web cleat connection. It shows

that the response of the component-based model is similar

to the experiments of the web cleat connection. However,

the maximum failure load and connection rotation of the

component-based model are slightly different from the

experimental test results. The maximum failure load in this

experiment was nearly two times higher than Yu’s (Yu

et al. 2009) test results. This clearly indicated that the tying

resistance of web cleat connection with stainless steel bolts

connection is higher than web cleat connection with

ordinary bolts. This also showed that the connection per-

formance and robustness at elevated temperature improved

when adopting stainless steel bolts in the connection.

In this connection, two types of failure mode were found

as follows: two web cleat angles underwent significant

amounts of deformation and the top two bolts connected to

the column flange were deformed as shown in Fig. 11. This

fracture happened due to a very high rotation of this con-

nection. This type of failure in the web cleat connection

depends on temperature and rotation of connection.

Prediction of the maximum failure load

The component-based model analysis of the model gives a

reasonable structural response to the fin plate connection

and web cleat connection. Based on the component-based

model analysis of the model, this study predicts the max-

imum failure load of different types of connections. This

study involved eight simple models using the fin plate

connection and eight simple models using the web cleat

connection. One of the assumptions made in this analysis is

that the temperature distribution is uniform. Those models

are based on the model specifications detailed in

Sects. 2.3.1 and 2.3.2. However, this model is separated

into two types: connection with ordinary bolts and with

stainless steel bolts. Each category has four different beam

sections and four different numbers of bolts’ rows. The

above-detailed model used the same column section, i.e.

254 9 254 9 89 UC S355. The component-based results

are compared with experimental results. Figures 12 and 13

show the comparison of the force–rotation relationship for

the fin plate connection and web cleat connection with the

experimental test results. This figures also indicated that

the component-based model results have given an accept-

able prediction of the test behaviour. Based on this, the

study predicts different connections maximum failure load

at elevated temperature as shown in Tables 1 and 2.

Conclusions

The response of the fin plate connection and web cleat

connection using stainless steel bolts to improve robustness

in fire has been studied using experimental tests and

component-based model analysis. The outcomes of this

study will be used to assess the vulnerability of the fin plate

connection and web cleat connection during fire hazards.

The experimental results show that the maximum failure

load of stainless steel bolted connections is higher than

connections with ordinary high strength (8.8) bolts. The

study indicates that the adoption of stainless steel bolts

connection could improve the robustness of the fin plate

and web angle connections in fire.

The component-based model gives a reasonable pre-

diction of connection performance with both 8.8 and

0102030405060708090

100110120130140

0 1 2 3 4 5 6 7 8 9 10 11 12

Forc

e (k

N)

Rotation (Degree)

Component based model for Stainless steel bolts

Stainless steel Test

Fig. 9 The comparison of force–rotation relationships of fin plate

connection at temperature 550 �C

0

10

20

30

40

50

60

0 5 10 15 20 25

Forc

e (k

N)

Rotation (Degree)

Component based model Stainless steel bolt

Stainless steel Test

Fig. 10 The comparison of force–rotation relationships of web cleat

connection temperature 650 �C


123

stainless bolts. The component-based design method is an

alternative method to assess the robustness of a connection

in design practice. A comparison was made of the potential

for improved connection performance if stainless steel

bolts were used. It is recommended based on this study and

earlier investigations:

• To improve the performance of conventional steel

connections, stainless steel bolts could be used to avoid

bolt shear failure and change the failure mode to web

bearing.

Fig. 11 The failure modes of

a Fin plate b Web cleat

connections

0102030405060708090

100110120130140

0 1 2 3 4 5 6 7 8 9 10 11 12

Forc

e (k

N)

Rotation (Degree)

Component based model for Ordinary boltsComponent based model for Stainless steel boltsTest (Yu [3])Stainless steel Test

Fig. 12 The comparison of force–rotation relationship of component-

based model for fin plate connection at 550 �C

0

10

20

30

40

50

60

0 5 10 15 20 25

Forc

e (k

N)

Rotation (Degree)

Component based model for Ordinary boltsComponent based model Stainless steel boltTest (Yu, [6])Stainless steel Test

Fig. 13 The comparison of force–rotation relationship of component-

based model for web cleat connection at 650 �C

Table 1 The prediction of maximum load for fin plate connection at

temperature 550 �C

Fin plate, double line of bolts

150 9 10 mm fin plate in S275

Beam size (UB,

S275)

Bolt rows, n F8.8

(kN)

Fss

(kN)

LF

(mm)

406 9 178 9 54 4 125.19 250.26 260

457 9 152 9 52 5 168.94 325.03 320

533 9 210 9 82 6 229.42 437.95 380

Table 2 The prediction of maximum load for web cleat connection

at temperature 650 �C

Double angle cleat, single line of bolts

2 No. 90 9 90 9 8 mm equal angle in S275

Beam size (UB, S275) Bolt rows, n F8.8 (kN) Fss (kN) la (mm)

305 9 165 9 40 3 24.74 47.22 200

406 9 178 9 54 4 36.80 65.89 260

457 9 152 9 52 5 37.79 84.13 320

533 9 210 9 82 6 62.84 121.68 380

F8.8 maximum force in ordinary bolt connection

Fss maximum force in stainless steel bolt connection

lF fin plate length

la angle cleat length


123

• The fin plate connection with stainless steel bolts would

need to be designed to avoid brittle failure models at

elevated temperature such as weld fracture.

Acknowledgments The authors gratefully acknowledge the assis-

tance provided by Dr Shan–Shan Huang in conducting the experi-

ments and the support of the technical staff in the Heavy Structures

Laboratory at the University of Sheffield.

Open Access This article is distributed under the terms of the

Creative Commons Attribution License which permits any use, dis-

tribution, and reproduction in any medium, provided the original

author(s) and the source are credited.

References

Al-Jabri KS, Davison JB, Burgess IW (2008) Performance of beam-

to-column joints in fire, A review. Fire Saf J 43(1):50–62

BSI (1990) BS5950: Part 1:1990 structural use of steelwork in

building—Part1: code of practice for design in simple and

continuous construction: hot rolled sections. British Standards

Institution, London

CEN (2002) BS EN 1990:2002 Eurocode 0: basis of structural design.

European Committee for Standardisation, Brussels

Jaspart JP, Demonceau JF (2008) European design recommendations

for simple joints in steel structures. J Constr Steel Res

64:822–832

Sarraj M, Davison JB, Burgess IW, Plank RJ (2007) Finite element

modelling of fin plate steel connections in fire. Fire Saf J

42:408–415

UFC 4-023-03 (2009) Unified Facility Criteria (UFC) design of

buildings to resist progressive collapse, Approved for public

release, Distribution unlimited pp12

Way AGJ (2011) Structural robustness of steel framed buildings.

Publication P391, The Steel Construction Institute, Ascot

Yu HX, Burgess IW, Davison JB, Plank RJ (2009a) Experimental

investigation of the behaviour of fin plate connections in fire.

J Constr Steel Res 65:723–736

Yu HX, Burgess IW, Davison JB, Plank RJ (2009b) Tying capacity of

web cleat connections in fire. Part 1: test and finite element

simulation. Eng Struct 31(3):651–663

Yu HX, Burgess IW, Davison JB, Plank RJ (2009c) Tying capacity of

web cleat connections in fire. Part 2: development of component-

based model. eng Struct 31(3):697–708


123

Robustness of steel joints with stainless steel bolts in ﬁre · Keywords Steel joints Stainless...

Documents

Transcript of Robustness of steel joints with stainless steel bolts in ﬁre · Keywords Steel joints Stainless...